Linear Copolymers of Ethylene and Polar Vinyl Monomers via Olefin

Macromolecules 2007, 40, 2643-2656

2643

Linear Copolymers of Ethylene and Polar Vinyl Monomers via Olefin Metathesis-Hydrogenation: Synthesis, Characterization, and Comparison to Branched Analogues Stephen E. Lehman, Jr.,†,§ Kenneth B. Wagener,*,† Lisa Saunders Baugh,*,‡ Steven P. Rucker,‡ Donald N. Schulz,‡ Manika Varma-Nair,‡ and Enock Berluche‡ The George and Josephine Butler Polymer Research Laboratory, Department of Chemistry, UniVersity of Florida, GainesVille, Florida 32611-7200, and ExxonMobil Research & Engineering Company, Corporate Strategic Research Laboratory, 1545 Route 22 East, Annandale, New Jersey 08801 ReceiVed January 11, 2007; ReVised Manuscript ReceiVed January 29, 2007 ABSTRACT: Materials structurally equivalent to copolymers of ethylene and 99% tert-butyl cyclooct4-enecarboxylate (8), which can be easily hydrolyzed to cyclooct-4-ene carboxylic acid (9).7b The isomeric mixture of methyl ester-substituted cyclooctenes 7/7a/7b was employed for copolymerization since the mixture actually provides a greater degree of sequence distribution randomness in the copolymer than the single isomer. However, simple base-catalyzed esteri-

Macromolecules, Vol. 40, No. 8, 2007

Figure 2. GC kinetic plots of monomer conversions for ROMP copolymerizations at 60 °C: (a) 5; (b) 6; (c) 7/7a/7b; (d) 8.

fication22 of 9 provided >99% isomerically pure 7, which allowed the identification of the isomer mixture components by 1H NMR and may offer a convenient route into a variety of useful 5-substituted cyclooctene ROMP monomers. ROMP was conducted on mixtures of cyclooctene and 5-9 using “second-generation” imidazole-based Ru catalysts to produce high molecular weight EVAC, EVOH, EMA, EAA, and ETBA copolymers having theoretical 2, 4, and 6 mol % polar vinyl monomer incorporations.26 Most of the polymerizations were carried out with RuCl2(SIMes)(PCy3)CHCH3 (Ru3), an ethylidene complex recently reported by the Wagener group,14 in order to install rigorously linear alkyl end groups on the polymers during initiation. A few samples were prepared with RuCl2(SIMes)(PCy3)CHdCH(CH3)2 (Ru-4).15 Most of the copolymerizations were conducted on an 8-14 g monomer scale at 60 °C in o-dichlorobenzene (ODCB), a good solvent for linear polyethylenes (providing convenience for tandem ROMPhydrogenation), at initial monomer concentrations of 0.4-1.1 M.27 In contrast to the moderate molecular weight materials obtained with ADMET, ROMP copolymerization gave copolymers with Mns of 45 000-185 000 g/mol (by GPC vs PE standards) after hydrogenation (Table 1). The kinetics of the ROMP copolymerizations are important in determining the microstructures of the resultant functional polyethylenes. Equal reactivity of the two comonomers is desired to produce materials that best approximate random copolymers of ethylene and functional vinyl monomers. Kinetics for the 6 mol % polar vinyl EVOH, EVAC, EMA, and ETBA copolymerizations were monitored by GC aliquot analysis (Figure 2). The 60 °C copolymerization kinetics of cyclooctene with alcohol-substituted 5 were found to approach completely random incorporation of the two monomers. However, tracking of the 60 °C copolymerizations involving ester- and acetate-functionalized 6-8 indicated that cyclooctene is polymerized at a slightly greater rate than these functional cyclooctenes. Thus, while the copolymerizations were complete within approximately 5 min at 60 °C, they were allowed to proceed for longer periods

Linear Ethylene/Polar Vinyl Copolymers 2649

(typically 2-3 h) in order to promote sequence distributionrandomizing olefin isomerization and cross-metathesis events. Model 4-6 mol % comonomer copolymerizations were also studied at room temperature by 1H NMR and/or GC for 5, 6, and 7/7a/7b. The relative rates of conversion between the functionalized and unfunctionalized cyclooctene monomers are not altered from those seen at 60 °C although the overall rate of polymerization decreases (total reaction time e2 h). Roomtemperature 1H NMR kinetics with the 7/7a/7b mixture revealed that the 5-substituted isomer 7 is incorporated faster than 4-substituted 7a, whereas the 3-substituted isomer 7b is not polymerized. Most of the ROMP polyalkenamers were hydrogenated by simply applying 400 psig H2 to the Ru-3- or Ru-4-containing polymerization solution at 130 °C. A similar procedure has been used by McLain et al. to hydrogenate functional cyclooctene ROMP homopolymers.6b,28 This in situ Ru-based hydrogenation procedure was satisfactory for the ester- and acetate-containing polymers, but resulted in some conversion of alcohol groups to ketones, as was observed for ADMET EVOHs hydrogenated using Ru-1. Ru-catalyzed double bond migration23 of a polyalkenamer olefin to form an enol followed by tautomerization to the corresponding ketone is a potential mechanism for this transformation (in addition to transfer hydrogenation).25a Hydrogenation of one EAA copolymer using Ru-3 resulted in conversion of some acid groups to secondary alcohols and ketones through an unknown mechanism, presubably via decarbonylation. To circumvent these issues, the remaining EAA polyalkenamers and an additional set of EVOH polyalkenamers were isolated and hydrogenated via diimide reduction using p-toluenesulfonhydrazide.9b Copolymers hydrogenated using this procedure were largely free of defect functionalities. Spectral Analysis of Linear Copolymers by NMR and IR. High-temperature 1H and 13C NMR compositional analysis for the model linear functional polyethylenes (reported as the equivalent “mol % ethylene” and “mol % polar vinyl comonomer,” Table 1) showed good agreement with compositions expected from monomer feed ratios. 13C NMR EAA analysis was complicated by low polymer solubility and the inability to use Cr(acac)3 relaxation agent (due to ligand exchange with the polymeric acid groups); higher-acid-content copolymers could only be analyzed by 1H NMR or solid-state 13C NMR. The 13C NMR spectra of these rigorously linear copolymers provide nonambiguous assignments for isolated functional polar vinyl comonomer units (EVE triads) in ethylene/polar vinyl copolymers, without the complexity of short-chain alkyl branching or polar vinyl dyads or triads (EVV or VVV units). The 13C NMR assignments for all of the observed polymer segments are given in Figure 3. Figure 4 shows a spectral comparison for a linear ROMP EVAC and an alkyl-branched free-radical EVAC of similar composition. The copolymer IR spectra showed two pairs of closely spaced bands at 1473/1463 and 730/720 cm-1, in addition to characteristic bands associated with functional groups. These features are consistent with the orthorhombic phase of polyethylene.29 EAA IR spectra indicated that the acid groups exist exclusively as hydrogen-bonded dimers in the amorphous phase at room temperature, similar to free-radical EAA copolymers (υCdO at 1706 cm-1 as opposed to 1750 cm-1 for free acid groups).30 Molecular Weight and Branch Analysis by GPC and GPC)LS. High-temperature GPC analysis using PE standards was used for copolymer molecular weight analysis. Although this technique is complicated by potentially significant, compositionally dependent differences in solution behavior between

2650

Lehman et al.

Figure 3. 13C NMR assignments for all observed segments in linear functional polyethylenes (120 °C, ODCB-d4 (EMA) or TCE-d2 (all others); the CHO2CMe peak for EVAC is not resolved from TCE-d2).

the PE standards and the functionalized materials, it indicated that the ROMP polymers were indeed of high molecular weight (Mn g 45 000). The EAA copolymers exhibited apparent molecular weights (vs PE) considerably lower than the other ROMP materials (Mn 12 000-21 000). Given their limited solubility and the fact that no end groups were observed by NMR, it is likely that the EAAs are actually of comparable molecular weight to the other copolymers. GPC-LS analysis of free-radical EVAC, EVOH, and EMA copolymers (using composition-specific input parameters) indicated that GPC underestimates Mn for these “milder” functional materials by a factor of approximately 2-3 (Table SI-2, Supporting Information); it is likely that this effect is more pronounced for the acid-containing materials. The effect of ROMP polymerization temperature on molecular weight is illustrated by the significantly higher Mns of the three EVOH samples prepared at room temperature (Mn 123 000-185 000 by GPC vs PE) as compared to the samples prepared at 60 °C (Mn 47 000-95 000). GPC-LS g′ calculations were also performed for selected ADMET EVOH and EVAC samples to probe long-chain branch (LCB) content and further confirm the linear nature of the metathesis materials. The g′ values for the ADMET copolymers ranged from 0.82 to 1.3331 and were typically at least twice as

Figure 4.

13C


high as g′ values for free-radical samples of similar composition. For example, an ADMET EVAC containing 2.1 mol % VAC showed a g′ value of 1.01, indicating linearity, while an identically composed (2.1 mol % VAC) free-radical copolymer had a g′ value of 0.45, indicating significant LCB (Figure SI-3, Supporting Information). Thermal Analysis by DSC. Substituent Effects on Tm and Xc. The fundamental effects of branches in polyethylenes are generally understood in terms of the Flory exclusion model, whereby branches of sufficient size (two or more non-hydrogen atoms) are excluded from the crystal lattice into the amorphous region.32 The effect of excluded groups on lowering crystallinity (Xc) and melting point (Tm) is purely colligative in nature, and is independent of branch length or chemical identity at low comonomer contents.33 Very small functionalities (Cl, OH, ) O, Me)34 are capable of partial inclusion in the polyethylene lattice on an equilibrium basis or as point defects according to the inclusion model of Sanchez and Eby.35 Groups of intermediate size (such as ethyl) may partially enter crystals to a small extent but more characteristically act as amorphous defects and exert a disordering influence on the lattice.36 The thermal properties of free-radical, alkyl-branched, LDPElike EVOH, EVAC, EMA, and EAA copolymers have been extensively studied.30a,33b-c,34b,37 Trends for EVAC, EMA, and EAA copolymers are consistent with the exclusion model and similar to those seen for ethylene/R-olefin copolymers; both Tm and Xc decrease rapidly as the amount of pendent functional groups increases.33c For example, free-radical EVAC copolymers become completely amorphous at VAC contents of g25 mol %.37c In contrast, the crystallinity and melting temperatures of free-radical EVOH copolymers are depressed to a lesser extent, due to the smaller size of the OH group and its ability to be partially incorporated into the polyethylene crystal lattice. The behavior of EVOH copolymers therefore more closely resembles that of ethylene/CO and EP copolymers than that of ethylene/ R-olefin copolymers. Sequence distribution effects on the thermal properties of branched and functional polyethylenes are also well-studied.32,33b,37a,38 For a given substituent content, polymers having a broad distribution of methylene run lengths (and therefore containing longer maximum methylene sequences) typically have higher Tms than polymers with more evenly distributed substituents. This is not only because their long methylene

NMR comparison of EVAC copolymers produced by ROMP/hydrogenation and free-radical polymerization.



Figure 5. Tm and crystallinity vs mol % comonomer for linear functional polyethylenes prepared by ROMP/hydrogenation (low CO e 0.1 mol % H2CdCdO units; high CO g 0.3 mol % H2CdCdO units). Table 2. Thermal and Mechanical Properties of Linear Functional Ethylene Copolymers Synthesized via Metathesis polymer

Tm(max), °C (Xc)a

EVOH-1.9-R EVOH-1.9-R-2 EVOH-2.0-R EVOH-3.4-R EVOH-3.5-R EVOH-3.6-R EVOH-5.3-R EVOH-6.1-R EVAC-1.9-R EVAC-3.3-R EVAC-4.6-R EMA-1.9-R EMA-3.6-R EMA-5.6-R ETBA-1.8-R ETBA-3.5-R ETBA-5.3-R EAA-2.6-R EAA-3.4-R EAA-8.3-R PE-R EVOH-2.2-A EVOH-2.3-A EVOH-4.7-A EVOH-5.1-A EVOH-6.1-A EVAC-2.1-A EVAC-6.2-A

128.8 (0.584) 127.8 (0.587) 127.4 (0.462) 126.5 (0.539) 125.3 (0.543) 125.1 (0.448) 123.7 (0.437) 120.5 (0.389) 109.8 (0.373) 97.9 (0.319)c 89.6 (0.284)c 111.8 (0.378)c 100.5 (0.322)c 85.4 (0.265)c 109.4 (0.344)c 96.6 (0.251)c 81.6 (0.246)c 114.9 (0.408) 103.3 (0.312) 89.0 (0.219)c 130.8 (0.564) 128.5 (0.705) 126.9 (0.830) 123.1 (0.770) 121.8 (0.676) 115.4 (0.659)c 117.8 (0.532) 73.2 (0.199)c

25 °C storage modulus, MPa (mode)b 687 (B) 552 (B) 666 (T) 677 (B) 650 (B) 672 (T) 565 (B) 684 (T) 204 (B)d 111 (B) 92.1 (B) 340 (B) 75.8 (B) 52.2 (B)f 221 (T) 87.9 (T) 49.1 (T) 831 (T) 463 (T) 142 (T) 967 (T) 1418 (B)

451 (B)

DSC. b DMTA; B ) 3-point bend deformation; T ) tensile deformation. c Broad or with low-temperature shoulder. d T ) 251 MPa. e T ) 70.3 MPa. a

sequences can crystallize to form relatively large (high-melting) lamellae, but bcause some portion of their substituents are in dyad or block configurations, behaving as one large defect rather than separate defects (lessening the impact per branch). At the other extreme, the content of crystallizable sequences is reduced for copolymers with evenly spaced substituents, and the shorter methylene sequences present generally form smaller lamellae and produce lower Tms. This effect is strikingly illustrated by the unusually low Tms (and/or unusually high Xcs and narrow melts due to greater uniformity of crystal size) seen for many “preciselybranched”modelADMETpolyethylenes.11b-d,f,i,j,l,n,o,12c,13 Comparison between Functional Groups for Linear Copolymers. DSC analysis was conducted on the linear ROMP copolymers to measure Tm (maximum) and Xc (Table 2); results are plotted in Figure 5. The ROMP copolymer materials, in contrast to previous functional ADMET models having “precise” sequence distributions, display broad, high-temperature Tms more characteristic of true random ethylene/polar vinyl copolymers. For all five comonomer types, both Tm and Xc decrease with increasing functional group content. Representative curves

Figure 6. Second heat DSC curves for linear ROMP ETBA copolymers with varying TBA contents. Table 3. Thermal and Mechanical Properties of Commercial Free-Radical Functional Ethylene Copolymers

polymer

comp (as mol % polar vinyl)a

Tm(max), °C (Xc)b

EVOH-2.4-FR EVOH-3.9-FR EVOH-5.8-FR EVAC-1.8-FR EVAC-1.9-FR EVAC-2.1-FR EVAC-3.0-FR EVAC-3.0-FR-2 EVAC-4.0-FR EVAC-5.3-FR EVAC-6.3-FR EMA-2.0-FR EMA-2.2-FR EMA-3.3-FR EMA-4.9-FR EMA-5.7-MA EMA-7.8-FR EMA-11.4-FR EAA-1.9-FR EAA-2.4-FR EAA-3.3-FR EAA-3.3-FR-2 EAA-3.7-FR EAA-3.8-FR EAA-5.5-FR EAA-6.4-FR

2.4 VOH 3.8 VOH, 0.1 ketene 5.8 VOH 1.7 VAC, 0.1 VOH 1.9 VAC 2.1 VAC 3.0 VAC 3.0 VAC 4.0 VAC 5.3 VAC 6.3 VAC 2.0 MA 2.2 MA 3.3 MA 4.9 MA 5.7 MA 7.8 MA 11.4 MA 1.9 AA 2.4 AA 3.3 AA 3.3 AA 3.7 AA 3.8 AA 5.5 AA 6.4 AA

107.0 (0.470) 107.6 (0.487) 107.9 (0.405) 102.9 (0.313) 102.5 (0.286) 101.4 (0.315) 96.2 (0.278) 98.8 (0.262) 92.2 (0.227) 88.6 (0.163) 84.9 (0.152) 100.8 (0.311) 101.1 (0.287) 99.7 (0.263) 88.3 (0.184) 86.4 (0.165) 75.1 (0.078) 64.6 (0.035) 97.5 (0.284) 102.5 (0.312) 95.1 (0.195)d 98.3 (0.255) 96.0 (0.234) 95.3 (0.231) 83.4 (0.099) 89.2 (0.141)

25 °C storage modulus, MPa (mode)c 361 (B) 182 (B) 209 (B) 82.5 (B) 78.1 (B) 56.7 (B) 64.6 (B) 32.5 (B) 28.1 (B) 26.6 (B) 140 (B) 22.2 (B) 26.6 (B) 14.9 (B) 12.0 (B) 344 (T) 196 (T) 193 (T)e 38.8 (T)

a Average of 1H and 13C NMR. b DSC; all T s except for EVOHs were m broad or exhibited a low-temperature shoulder. c DMTA; B ) 3-point bend deformation; T ) tensile deformation. d Two maxima, 92.0 and 98.1 °C. e B ) 158 MPa.

for ETBA copolymers are shown in Figure 6. As predicted by the Flory exclusion model and literature data for other branched polyethylenes, the identity of the functional groups in the EVAC,

2652

Lehman et al.


Figure 7. Tm and crystallinity vs mol % comonomer for alkyl-branched free-radical functional polyethylenes.

Figure 8. Tm (top) and crystallinity (bottom) comparison for linear (methathesis/hydrogenation) vs alkyl-branched (free-radical) functional polyethylenes (low CO e 0.1 mol % H2CdCdO units; high CO g 0.3 mol % H2CdCdO units). Comparative “precisely branched” homoADMET EVAC data taken from ref 11b.

EMA, and ETBA copolymers is irrelevant. Tm and Xc for these three series vary only with comonomer content. Compositional uncertainties for the ROMP EAA copolymers complicate a determination of whether the acid group behaves similarly: two of the EAA copolymers (EAA-2.6-R and EAA-8.3-R) exhibited Tms and Xcs higher than those for the ester/acetatesubstituted copolymers. This may potentially be explained by the presence of secondary alcohol groups and ketones in EAA2.6-R and uncertainty from use of a different compositional analysis method (solid-state 13C NMR) for EAA-8.3-R. No significant Tm and Xc differences are observed between acidand ester/acetate-substituted free-radically prepared copolymers (vide infra). Consistent with partial substituent inclusion into the crystalline phase,33b the melting temperatures of the EVOH metathesis copolymers were much higher (and narrower) than those for

the acid- and ester/acetate-functionalized copolymers, and showed a gentler depression in Tm per comonomer unit. Partial replacement of the OH substituents with ketones (“high CO” samples, Figure 5) did not affect Tm but increased Xc and the sensitivity of Xc to composition. It is possible that the lower Xcs seen for the CO-free EVOHs as compared to the “high CO” samples are a reflection of their higher molecular weights.33e,39 Effects of Alkyl Branching and Molecular Weight. A suite of low-polar-content free-radical EVAC, EMA, and EAA copolymers, containing 10-43 short-chain alkyl branches per 1000 total carbons (Table SI-2, Supporting Information), was studied for comparison to the linear ROMP materials. Comparative alkyl-branched EVOH copolymers were prepared by hydrolysis of branched EVACs. DSC Tm and Xc data for the free-radical materials (Table 3) are plotted in Figure 7. DSC curve shapes and general trends parallel those seen for the linear



Figure 9. Room-temperature storage moduli for (a) linear (ROMP/hydrogenation) and (b) alkyl-branched (free-radical) functional polyethylenes (low CO ) e0.1 mol % H2CdCdO units; high CO ) g0.3 mol % H2CdCdO units). B ) three-point bend mode; T ) tensile mode (*ROMP ) T, FR ) B).

Figure 10. Functional group effects on modulus for linear functional polyethylenes having similar composition (left to right): EMA-3.6R, EVAC-3.3-R, EVOH-3.5-R (0.3 mol % H2CdCdO), EVOH3.6-R, EAA-3.4-R, and ETBA-3.5-R.

copolymers: the branched ester/acetate-substituted materials exhibit colinear depressions in Tm and Xc with no functionalityspecific effects, while the branched EVOHs show comparatively higher and narrower Tms, greater Xcs, and a reduced sensitivity of both to composition. Overall, the branched EAA materials do not show behavioral differences from the ester/acetatesubstituted copolymers. Tm and Xc comparisons between each set of linear and alkylbranched copolymers are given in Figure 8. The linear esterand acetate-substituted ROMP copolymers show higher Tms than the branched copolymers at the lowest comonomer contents (below ca. 4 mol %), although the differences are small (Figure 8, top). The linear and branched Tm vs composition curves converge at ca. 4-7 mol % comonomer. Compositional uncertainties for samples EAA-2.6-R and EAA-8.3-R obscure trends for the acid-substituted copolymers, although the data is suggestive of a similar relationship. In contrast, the linear ROMP EVOHs showed higher Tms than the alkyl-branched copolymers at all compositions, consistent with partial OH inclusion into the crystal lattice and exclusion of the free-radical copolymer alkyl branches. The ADMET EVAC and EVOH materials (also plotted in Figure 8) had Tms similar to those for the higher-Mn ROMP materials. This data point suggests that at comonomer contents of greater than ∼7 mol %, linear functional copolymers with larger “excluded” substituents will have equal or lower Tms than alkylbranched materials, despite the fact that the linear polymers contain fewer overall chain punctuations. Linear materials with

>7 mol % randomly distributed polar comonomer were not available to probe this regime. However, assuming an extension of the linear relationship between alkyl-branched copolymer Tm and mol % VAC into this compositional range as a lower boundary,33c,40 the Tms for “precisely” branched model ADMET EVACs having 7.4-10.5 mol % VAC11b do indeed fall below those predicted for alkyl-branched samples (Figure 8, top left). On the basis of the “leveling” effect of alkyl branches and the greater run length-shortening effect of more evenly distributed substituents, “precisely” functionalized substituents would be expected to show the greatest per-unit lowering effect on Tm, with functional groups on alkyl-branched copolymers showing the weakest effect, and “random” linear materials intermediate between the two. The linear copolymers with “excluded” acid and ester/acetate substituents showed higher Xcs than the branched copolymers at all compositions studied (Figure 8, bottom). Interestingly, while the “high CO” ROMP EVOH materials had, on average, higher Xcs than the alkyl-branched EVOHs, the CO-free ROMP EVOHs were not demonstrably more crystalline than the branched EVOHs despite their linearity. Again, this may reflect the higher molecular weights of the three CO-free EVOH samples; crystallinities for the all of the low-Mn ADMET materials (both high- and low-CO EVOH and EVAC) were decidedly higher than those for analogous ROMP samples. Thermomechanical Analysis by DMTA. While Tm and Xc are functions of a material’s crystalline fraction, the storage modulus of a semicrystalline polymer reflects both its crystallinity and the cross-link density of its amorphous phase. Moduli measurements are therefore potentally more sensitive to chemical differences between functional substituents (even those excluded from the crystalline phase). Relationships between composition and mechanical properties for alkyl-branched freeradical EVAC and EVOH copolymers have been studied by Nielsen37d and Salyer and Kenyon.37c For EVACs, shear modulus, Clash-Berg modulus, and tensile strength were found to decrease with increasing VAC content. However, since mechanical properties are particularly sensitive to branch identity and content, trends observed for highly branched functional polyethylenes may not be predictive for linear materials. Temperature-dependent DMTA modulus data were collected on compression-molded linear and free-radical functional copolymer samples using either tensile or three-point bend deformation (Tables 2 and 3).41 Plots of 25 °C storage moduli vs comonomer content are shown in Figure 9. Moduli ranged between 10 and 700 MPa, as compared to 967 MPa for a perfectly linear unfunctionalized polyethylene prepared by ROMP/hydrogenation (the low-Mn ADMET materials showed

2654

Lehman et al.


Figure 11. Storage modulus comparison for linear (ROMP/hydrogenation) vs alkyl-branched (free-radical) functional polyethylenes (low CO e 0.1 mol % H2CdCdO units; high CO g 0.3 mol % H2CdCdO units). B ) three-point bend mode; T ) tensile mode.

Figure 12. Three-point bend mode DMTA curves for linear and branched EMA copolymers (25 °C modulus values in parentheses).

anomalously high moduli of 450-1420 MPa, 1.5-2.5 times greater than ROMP materials of similar composition, and are not included in Figure 9). For both linear and alkyl-branched ester/acetate-substituted copolymers, 25 °C storage modulus decreases with increasing functional group content. Modulus values are not sensitive to the identity of the functional substituent except possibly at the very lowest comonomer contents, where moduli for EMAs were higher than for EVACs and ETBAs. This general trend is consistent with a colligative crystallinity-depressing effect involving excluded, weakly interacting substituent groups. Modulus also decreases with increasing acid content for the EAA copolymers (reflecting a dominating factor of loss of crystallinity), but values are higher than for the analogous ester/acetate-substituted copolymers, presumably as a result of amorphous-phase hydrogen bonding between acid groups. Overall, the OH-substituted copolymers had the highest moduli, consistent with the alcohol group’s ability to both enter the crystal and potentially participate in amorphous cross-links (vide infra). This effect was most pronounced for the linear EVOH copolymers, which, in contrast to the alkyl-branched materials, showed no apparent sensitivity to composition.42 A summary comparison of functional group effects for linear copolymers having ca. 3.5 mol % polar comononer is given in Figure 10. Storage modulus comparisons for each linear vs alkylbranched copolymer set are given in Figure 11; representative DMTA curves for EVAC materials are shown in Figure 12.

The ester/acetate- and acid-substituted copolymers showed relationships very similar to those exhibited for Tm: the relatively higher moduli of the linear materials at low comonomer content appeared to trend toward convergence with those for alkyl-branched materials at ca. 5-6 mol % comonomer. By contrast (and also analogous to Tm trends), moduli for the linear EVOH copolymers were higher than those for the alkylbranched copolymers at all studied compositions. The unusual insensitivity of modulus to VOH content for the linear EVOH polymers is interesting. The OH substituents in alkyl-branched EVOH materials are known to associate via hydrogen bonding.37c,43 It is likely that removing the alkyl branches in EVOHs allows for more efficient hydrogen bonding interactions between OH groups in the amorphous phase, and, unlike for the excluded acid groups in EAAs, possibly also in the crystalline phase. This may explain the superior Tms, crystallinities, and storage moduli observed for these materials. Conclusions Alkyl-branch-free polyethylenes were synthesized by ruthenium-catalyzed olefin metathesis copolymerization followed by polymer hydrogenation and were shown to be structurally equivalent to copolymers of ethylene and 0.9 M precluded facile handling and transfer to hydrogenation vessels. (28) Other tandem Ru-based polyalkenamer hydrogenations are known: (a) Drouin, S. D.; Yap, G. P. A.; Fogg, D. E. Inorg. Chem. 2000, 39, 5412-5414. (b) Drouin, S. D.; Zamanian, F.; Fogg, D. E. Organometallics 2001, 20, 5495-5497. (c) Fogg, D. E.; Amoroso, D.; Drouin, S. D.; Snelgrove, J.; Zamanian, F. J. Mol. Catal. A: Chem. 2002, 190, 177-184. (29) (a) Tashiro, K.; Sasaki, S.; Kobayashi, M. Macromolecules 1996, 29, 7460-7469. (b) Ungar, G.; Zeng, X.-B. Chem. ReV. 2001, 101, 41574188. (30) (a) Otocka, E. P.; Kwei, T. K. Macromolecules 1968, 1, 244-249. (b) Kang, N.; Xu, Y.-Z.; Wu, J.-G.; Feng, W.; Weng, S.-F.; Xu, D.-F. Phys. Chem. Chem. Phys. 2000, 2, 3627-3630. (31) The g′ values given for the functional copolymers were calculated using the Mark-Houwink parameters for PE and are therefore only approximate (giving rise to values distributed somewhat above and below the theoretical linear value of 1.0), since Mark-Houwink parameters can vary with copolymer composition. However, for materials with similar compositions, relatiVe g′ comparisons may be used to accurately characterize relative differences in LCB. (32) (a) Flory, P. J. J. Chem. Phys. 1949, 17, 223-240. (b) Flory, J. P. Trans. Faraday Soc. 1955, 51, 848-857. (33) (a) Isasi, J. R.; Haigh, J. A.; Graham, J. T.; Mandelkern, L.; Alamo, R. G. Polymer 2000, 41, 8813-8823. (b) Alamo, R. G.; Mandelkern, L. Thermochim. Acta 1994, 238, 155-201 and references therein. (c) Alamo, R. G.; Domszy, R.; Mandelkern, L. J. Phys. Chem. 1984, 88, 6587-6595. (d) Alamo, R. G.; Mandelkern, L. Macromolecules 1989, 22, 1273-1277. (e) Alamo, R. G.; Viers, B. D.; Mandelkern, L. Macromolecules 1993, 26, 5740-5747. (f) Alizadeh, A.; Richardson, L.; Xu, J.; McCartney, S.; Marand, H.; Cheung, Y. W.; Chum, S. Macromolecules 1999, 32, 6221-6235. (34) (a) Gomez, M. A.; Tonelli, A. E.; Lovinger, A. J.; Schilling, F. C.; Cozine, M. H.; Davis, D. D. Macromolecules 1989, 22, 4441-4451.


(35) (36)

(37)

(38) (39) (40) (41)

(42)

(43)

(b) Bodily, D.; Wunderlich, B. J. Polym. Sci., Part A-2: Polym. Phys. 1966, 4, 25-40. (c) Ke, B. J. Polym. Sci. 1962, 61, 47-59. (d) Richardson, M. J.; Flory, P. J.; Jackson, J. B. Polymer 1963, 4, 221236. (e) Pe´rez, E.; VanderHart, D. L. J. Polym. Sci., Part B: Polym. Phys. 1987, 225, 1637-1653. (f) Wunderlich, B.; Poland, D. J. Polym. Sci., Part A 1963, 1, 357-372. (g) Baker, C. H.; Mandelkern, L. Polymer 1966, 7, 7-21. (h) Phillips, P. J.; Wilkes, G. L.; Delf, B. W.; Stein, R. S. J. Polym. Sci., Part A-2: Polym. Phys. 1971, 9, 499515. Sanchez, I. B.; Eby, R. K. J. Res. Natl. Bur. Stand. (U.S.) 1973, 77A, 353-358. (a) VanderHart, D. L.; Pe´rez, E. Macromolecules 1986, 19, 19021909. (b) Kawai, T.; Ujihara, K.; Maeda, H. Makromol. Chem. 1970, 132, 87-111. (c) Holdsworth, P. J.; Keller, A. J. Polym. Sci., Part B: Polym. Lett. 1967, 5, 605-612. (a) Keating, M. Y. Thermochim. Acta 1994, 243, 129-145. (b) Okui, N.; Kawai, T. Makromol. Chem. 1972, 154, 161-176. (c) Salyer, I. O.; Kenyon, A. S. J. Polym. Sci., Part A-1: Polym. Chem. 1971, 9, 3083-3103. (d) Nielsen, L. E. J. Polym. Sci. 1960, 42, 357-366. (e) Chowdhury, F.; Haigh, J. A.; Mandelkern, L.; Alamo, R. G. Polym. Bull. (Berlin) 1998, 41, 463-470. (f) Buerger, D. E.; Boyd, R. H. Macromolecules 1989, 22, 2699-2705. (g) Kortleve, G.; Tuijnman, C. A. F.; Vonk, C. G. J. Polym. Sci., Part A-2: Polym. Phys. 1972, 10, 123-131. (h) Vonk, C. G. J. Polym. Sci., Part C: Polym. Symp. 1972, 38, 429-435. (i) Takahashi, M.; Tashiro, K.; Amiya, S. Macromolecules 1999, 32, 5860-5871. (j) Buerger, D. E.; Boyd, R. H. Macromolecules 1989, 22, 2694-2699. (k) Otocka, E. P.; Kwei, T. K.; Salovey, R. Makromol. Chem. 1969, 129, 144-156. (l) MacKnight, W. J.; Taggart, W. P.; McKenna, L. J. Polym. Sci., Part C: Polym. Symp. 1974, 46, 83-96. (m) Horrion, J.; Agarwal, P. K. Polym. Commun. 1989, 30, 264-267. (n) Somrang, N.; Nithitanakul, M.; Grady, B. P.; Supaphol, P. Eur. Polym. J. 2004, 40, 829-838. (a) Chou, R. T.; Keating, M. Y.; Hughes, L. J. Annu. Tech. Conf.Soc. Plast. Eng. 2002, 60th (2), 1832-1836. (b) Gutzler, F.; Wegner, G. Colloid Polym. Sci. 1980, 258, 776-786. For the high-CO EVOH materials, it is also possible that a change in the heat of fusion for a perfect crystal from the PE value of 293 J/g towards the value for poly(ethylene-alt-CO) distorts the Xc calculation. Chen, H. Y.; Chum, S. P.; Hiltner, A.; Baer, E. J. Polym. Sci., Part B: Polym. Phys. 2001, 39, 1578-1593. In the discussion of DMTA data, some functional group comparisons are made using data obtained in different modes. Although their deformation processes differ (for example, elongation during tensile stress could potentially facilitate functional group interactions), modulus measurements for the two series of ROMP EVOHs (which were taken in different modes) showed reasonably good comparability, as did specific comparative measurements on selected other samples (Tables 2 and 3). Experimental limitations, including compositional uncertainties for ROMP EAAs and data scatter for the high-melting EVOHs, complicate any close examination of the relative behaviors of the EAA and EVOH copolymers. ROMP EVOHs appeared to have generally higher moduli than ROMP EAAs at all compositions (noting sample EAA-2.6-R contains ketone and alcohol groups), while free-radical EVOHs showed similar moduli to free-radical EAAs at low comonomer contents and higher moduli at higher comonomer contents. Kanekiyo, M.; Kobayashi, M.; Ando, I.; Kurosu, H.; Amiya, S. Polymer 2000, 41, 2391-2404.

MA070085P

Linear Copolymers of Ethylene and Polar Vinyl Monomers via Olefin

Recommend Documents